Over the past year, experts and researchers have raised alarms about AI’s voracious energy consumption. AI will contribute to a spike in power demand in the Middle East that the region hasn’t seen in decades.

Data centers operating constantly on full blast, and energy-hungry advanced AI chips are behind this growth.

INCREASED ENERGY DEMAND

“AI is highly energy-intensive. For example, a single Google search consumes approximately 0.3 watt-hours of electricity, a request to OpenAI’s ChatGPT requires  2.9 watt-hours—nearly ten times more,” says Jeremy Crane, Group CEO of Yellow Door Energy.  “This increased demand could potentially lead to a 60% rise in the energy requirements of data centers. The critical question is: where will this electricity come from?” 

In January, the International Energy Agency forecast that global data center electricity demand will double from 2022 to 2026, with AI playing a major role in that increase.

The Middle East’s still-exploding boom in AI could drive that number up a lot—and fast. 

The region’s tech industry’s hard pivot into AI has dramatically escalated the demand for data centers. Training AI models is extremely energy intensive and needs far more power  than traditional data center activities.

“AI’s energy consumption stems from data centers, which handle the massive computational tasks involved,” says James Thomas, Partner, Strategy& Middle East, part of the PwC Network. “Firstly, there is the training of AI models, especially in deep learning, which demands huge processing power. Then, there is inference – the computational processing needed to apply these models in response to each prompt. For example, a typical GPT prompt response requires around 10 times as much power (~3Wh) as a regular internet search (0.3Wh).” 

This energy usage has been exacerbated by the stiff competition between major tech companies racing to build more powerful AI models. Researchers say the cost of the computational power required to train these models is increasing, with no slowdown in sight.

A continuation of the current trends in AI capacity and adoption is set to lead to Nvidia shipping AI server units to the Middle East. These servers, running at full capacity, consume massive amounts of electricity. Recently, Nvidia has signed a deal to deploy its AI tech at data centers owned by Qatari telecoms group Ooredoo in Qatar, Oman, and Kuwait. Providing the technology will allow Ooredoo to help its customers better deploy Generative AI applications. Saudi Arabia is in talks to build a $18 billion data center with 1GW capacity at the Oxagon, Neom. 

“Countries in the region will continue to invest heavily in AI within the next few years,” says Abdulmajeed Harmoodi, VP of Engineering & Development at Khazna Data Centers, citing the IDC forecast on AI spending in the Middle East, Türkiye, and Africa region to soar at a five-year CAGR of 37%, with investments set to reach a monumental $7.2 billion in 2026.

SCALING CLEAN ENERGY

Although the region has a huge capacity to absorb the increased demand, many worry that the aggressive consumption of fossil fuel energy sources will hamper collective efforts to combat climate change. 

Experts say AI’s huge demand for power represents an opportunity for the countries here to scale the next generation of clean energy technologies and address energy efficiency.

“Scaling the clean-energy supply is key to meeting the rising power demands. Clean energy sources like solar and wind will also help cut down greenhouse gas emissions and fight climate change,” says Harmoodi.

In the Middle East, solar is the lowest-cost source of power during the day, and when coupled with the decreasing costs of the battery energy storage system (BESS), it will soon be competitive 24/7, adds Crane. 

“Rather than relying on fossil fuels or hydrocarbons, we can leverage  clean energy sources to meet the growing power demands of AI, ensuring a resilient and sustainable path forward.”

Although scaling clean energy supply is a critical response to the forecast spike in power demand driven by AI investments, Thomas says it isn’t the “only solution.” In parallel, he adds, it’s a chance for policymakers to address energy efficiency, energy storage, and diversification of energy sources to balance energy supply and demand.

Scaling clean-energy supply must be paired with optimizing energy efficiency, according to Renaud Capris, CEO of Enova. “AI has immense potential to enhance energy management through predictive analytics and smart grid systems. By integrating AI with renewable energy sources, Middle Eastern countries can meet growing demands while maintaining sustainability goals.”

CURRENT ENERGY INFRASTRUCTURE

The current energy infrastructure is insufficient to fully support the increasing demands of AI, the expected rise in electric vehicle usage, and 

increasing electrification of industrial processes, says Thomas.

“To meet these demands, substantial investments in renewable energy sources, energy storage, and grid modernization will be essential. In addition, innovation—in lower-energy AI technologies, for example— will help ensure a sustainable and resilient energy infrastructure for the future.”

Countries in the region are leading the way in the energy transition by spurring the construction of renewable energy sources, particularly wind, solar, and nuclear, and implementing renewable energy policies to encourage the private sector. 

However, there is room for improvement, says Crane. “As a solar developer working with businesses in the commercial and industrial sector, we face challenges such as caps on renewable energy projects and electricity tariff subsidies, which can limit the adoption of clean energy  solutions.”

“However, recent regulatory developments have been encouraging,” he adds.

At the moment, Harmoodi says the region needs significant investments in expanding and upgrading energy systems, integrating more renewable sources, and improving energy efficiency. 

In the region, many new data centers are still powered by fossil fuels, especially because it takes a while for new energy sources to come online. Also, many renewables don’t meet the needs of high-speed computing in their demand for consistent and high-quality energy.

Many say that if we ignore this opportunity, the Middle East will be disadvantaged in the race for the future of AI and energy production.

“Countries that don’t invest in clean energy technologies risk falling behind,” says Thomas. “Lagging behind in this area will hamper technological growth, competitiveness, and energy security.”

Crane says countries must proactively plan for the increased energy demand from AI usage. “Across the globe,  workers are quickly learning how to use AI in their everyday operations, underscoring the  urgency of preparing for this shift.”

He adds that by investing in clean energy technologies now, companies and countries can ensure they have the infrastructure needed to support AI energy usage and ensure competitiveness in this rapidly changing landscape. 

There are no two ways about this—the reflexive response to meeting AI’s growing power demands is a commitment to clean energy. The unprecedented scale and density of power needed for AI require a novel set of clean energy generation solutions, able to deliver reliable power 24-7 in ever-increasing amounts.

“The future of AI and energy is deeply interconnected, with clean energy providing the sustainable foundation for AI’s energy-intensive growth,” adds Capris. “Investing today ensures a competitive edge in tomorrow’s global energy and AI landscapes.”